A study on internal flow features and air entrainment effects in hydraulic jumps using numerical modeling techniques.

摘要

In spite of the general impression that the hydraulic jump is a well-known flow phenomenon, detailed theoretical and numerical models of the internal flow features in hydraulic jumps have yet to be developed. The purpose of this thesis is to model internal flow features as well as air entrainment in hydraulic jump through numerical means, in two dimensions, using two-phase flow theory. In particular, this paper stems off the numerical works already started by Gonzalez and Bombardelli (2005) in which the authors replicated the experimental works of Liu. et. al. (2004) through a sluice gate and flume setup. Two-phase flow models were executed using a state-of-the-art code that incorporates air entrainment at the free surface.The commercial code FLOW-3DRTM uses computational fluid dynamics (CFD) in which standard flow equations are discretized and solved in each user-defined cell. The mathematical model equations are solved by the method of finite volumes/finite differences in Cartesian space. Different numerical models were run with different input conditions and setups. First, a replication of the works by Gonzalez and Bombardelli (2005) yielded very similar data in terms of horizontal velocity profiles and total kinetic energy profiles. Next, the location of the upstream boundary condition was moved farther upstream and no significant effect was observed.Three other implementations were performed to account for a dip in water elevation directly behind the sluice gate. The first was a raised sluice gate that yielded a jump toe elevation at the height of the original simulation setup. Next, a fixed uniform velocity boundary condition was implemented to assure only initial movement in the horizontal direction. Lastly, a streamlined lip as presented by Liu et. al. (2004) in their experimental works was applied to the setup.It was observed that hydraulic jump flow characteristics differed from experimental data when steering from the original sluice gate setup. This is likely due to an unclearness of the experimental setup, namely intrusive measuring devices and a lack of error measurements presented in the literature. These simulations were also only run for a relatively low Froude number of 2.